A Zhejiang University team hooked 48 college students into 64-channel EEG caps to find the brain signature of TikTok addiction. Published June 27, 2024 in Frontiers in Human Neuroscience, the study measured a 0.395 correlation between higher addiction scores and suppressed frontal-lobe theta-wave activity during attention tasks.
Key facts
- 48 college students, 45 completed EEG recording
- Correlation minus 0.395 between addiction score and frontal theta spike
- P-value 0.007 for frontal-lobe effect
- Effect absent in parietal (minus 0.014) and central (minus 0.229) regions
- Self-control scale correlation minus 0.320, p=0.026
The experiment by Yan, Su, Xue, Hu, and Zhou ran each participant through the Attention Network Test: 192 trials of arrow-direction tasks measuring executive control. Before the task, participants completed a short-video addiction questionnaire scoring between roughly 25 and 75 (average 50.77).
The key finding: theta waves in the 4-8 Hz band, which normally spike in the frontal lobe during cognitive conflict, decreased as addiction scores rose. The correlation was minus 0.395 (p=0.007), and the effect persisted after controlling for anxiety, depression, age, and gender. Crucially, the signal was absent in parietal (minus 0.014) and central (minus 0.229) regions — only the frontal lobe showed suppression.
The unique take: This isn't just "social media is bad for attention." It's a direct electrophysiological measurement showing that the algorithmic recommendation system — a machine learning model optimized solely for engagement — systematically dampens the brain's executive-control circuits. The frontal lobe, responsible for self-control and focus, literally produces less of its characteristic conflict-resolution signal in heavy users. The authors' verbatim conclusion: "prolonged consumption of such content may primarily engage the lower-order cortical brain regions, such as those associated with emotional processing, and suppress activity in higher-order areas responsible for self-control and attention."
The same participants scored lower on a self-control scale (correlation minus 0.320, p=0.026), providing behavioral corroboration. The study design, while small (45 completions), is methodologically clean: a within-subjects EEG paradigm with validated behavioral metrics, not a correlational survey.
What to watch
Replication studies with larger, more diverse samples (not just college students) and longitudinal designs tracking theta-wave changes over months of platform use. Watch also for platform-specific comparisons: do TikTok's recommendation algorithm produce stronger suppression than Instagram Reels or YouTube Shorts?
